This relates generally to fuse sensing.
Fuse sensing refers generically to the sensing of memory arrays, including bits arranged in rows and columns. Generally, a row may be sensed at a time and the address for that row may be chosen in a number of different ways.
Examples of storage elements that use fuse sensing include fuses, non-volatile memories, one time programmable memories, and on-die programmable fuses. One characteristic of current fuse sensing protocols is that the fuse sensing is performed in the same order every time. The order or sequence of selected rows is always the same.
This exposes the fuse sensing protocol to so-called side channel attack methods, such as Differential Electromagnetic Analysis (DEMA), Differential Power Analysis (DPA), and Latent Trait Models (LTM), as examples. The values of the fuses can be determined during fuse sensing. By placing the device in an infinite reset loop and taking measurements over the same location of the device at the same time from the reset of each loop, a hacker can build up a signature of the fuse values within the device. As sensitive assets are sometimes stored in fuses, protection against these types of attacks is very important.